# US 2023/0267019 A1 · Perceptive scaling of computing resources

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2023/0267019 A1
- Document ID: US20230267019A1
- Status: Published application
- Kind: A1
- Application number: 17/676,653
- Filed: 2022-02-21
- Published: 2023-08-24
- Earliest claimed: 2022-02-21
- Assignees: Capital One Services, LLC
- Applicants: Capital One Services, LLC
- Inventors: Joshua Edwards; Michael Mossoba; Abdelkader M'Hamed Benkreira
- Web page: https://patentgrep.com/patent/US20230267019A1

## CPC

- G06F 9/5005: {to service a request}
- G06F 9/542: {Event management; Broadcasting; Multicasting; Notifications}
- G06F 9/505: {considering the load}
- G06N 3/08: Learning methods
- G06N 7/01: Probabilistic graphical models, e.g. probabilistic networks
- G06N 20/10: using kernel methods, e.g. support vector machines [SVM]
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS

## Abstract

Perceptive scaling of computing resources is disclosed. News sources can be analyzed to identify a future event as well as an occurrence date and type of the event. A computing resource affected by the event can be determined based on the type of event. A scaling strategy can be formulated for the computing resource, such as scaling up or out, to address resource load introduced by the event. The extent of scaling can be determined based on the predicted impact of the event on the computing resource. Subsequently, the scaling strategy can be scheduled for execution before the occurrence date. The scaling strategy can later be terminated or deactivated after the event terminates.

## Claims

### Claim 1

A perceptive resource scaling system, comprising:
a processor coupled to a memory that includes instructions that, when executed by the processor, cause the processor to:
invoke machine-learning-based text analysis to identify a future event and date of the future event from an electronic news source;
classify the event as a type of event based on text analysis of context surrounding the event in the electronic news source;
identify a computing resource affected by the event based on the type of event and a metatag associated with the computing resource that identifies a type of event that affects the computing resource;
generate a scaling strategy for the computing resource that accounts for a change in resource utilization associated with the event; and
trigger execution of the scaling strategy at a predetermined time before the date.

### Claim 2 (depends on 1)

The system of claim 1, wherein the instructions further cause the processor to:
infer impact of the event on the computing resource; and
generate the scaling strategy proportionate to the impact.

### Claim 3 (depends on 2)

The system of claim 2, wherein the instructions further cause the processor to infer the impact based on a number of news reports of the event.

### Claim 4 (depends on 3)

The system of claim 3, wherein the instructions further cause the processor to invoke a machine learning model that infers the impact based on similarity of the event to a prior event.

### Claim 5 (depends on 4)

The system of claim 4, wherein the scaling strategy comprises addition of computing resources.

### Claim 6 (depends on 1)

The system of claim 1, wherein the instructions further cause the processor to:
determine a duration of the event; and
schedule termination of execution of the scaling strategy at a predetermined time after the duration of the event.

### Claim 7 (depends on 1)

The system of claim 1, wherein the instructions further cause the processor to:
determine a confidence level associated with occurrence of the event; and
generate the scaling strategy based on the confidence level.

### Claim 8 (depends on 7)

The system of claim 7, wherein the instructions further cause the processor to determine the confidence level based on a number of news sources reporting the event.

### Claim 9 (depends on 1)

The system of claim 1, wherein the instructions further cause the processor to:
identify a dependent computing resource of the computing resource; and
generate the scaling strategy including the dependent computing resource.

### Claim 10

A method of perceptive scaling of computing resources, comprising:
executing, on a processor, instructions that cause the processor to perform operations associated with perceptive resource scaling, the operations comprising:
invoking machine-learning-based text analysis to identify a future event and date of occurrence from an electronic news source;
determining an event type based on analysis of context surrounding the event in the electronic news source;
identifying a computing resource impacted by the event based on the event type and metatag associated with the computing resource that identifies the event type as affecting the computing resource;
generating a scaling strategy for the computing resource; and
triggering execution of the scaling strategy a predetermined time before the date.

### Claim 11 (depends on 10)

The method of claim 10, the operations further comprising:
evaluating metatags associated with a plurality of computing resources; and
identifying the computing resource from the plurality of computing resources based on a match between the event type determined and the event type specified in the metatag of the computing resource.

### Claim 12 (depends on 10)

The method of claim 10, the operations further comprising:
inferring an impact of the event on the computing resource; and
generating the scaling strategy proportionate to the impact.

### Claim 13 (depends on 12)

The method of claim 12, the operations further comprising determining the impact based on a number of reports of the event.

### Claim 14 (depends on 12)

The method of claim 12, the operations further comprising inferring the impact based on comparison to a similar prior event.

### Claim 15 (depends on 12)

The method of claim 12, the operations further comprising inferring the impact based on information specified in the electronic news source.

### Claim 16 (depends on 10)

The method of claim 10, the operations further comprising:
identifying a dependent computing resource of the computing resource specified in the metatag; and
generating the scaling strategy includes scaling of the dependent computing resource.

### Claim 17 (depends on 10)

The method of claim 10, the operations further comprising:
determining a duration of the event; and
scheduling termination of the scaling strategy at a predetermined time after the duration of the event.

### Claim 18

A computer-implemented resource scaling method, comprising:
identifying a previously unknown future event and date of occurrence from an electronic news source with automated text analysis;
determining a type of the event based on analysis of context surrounding the future event in the electronic news source;
identifying a computing resource affected by the event based on the type of the event and metatags associated with computing resources that indicate event types that impact a computing resource;
generating a scaling strategy for the computing resource, wherein the scaling strategy adds computing resources to support an additional load from the event; and
triggering execution of the scaling strategy at a predetermined time before the date of occurrence.

### Claim 19 (depends on 17)

The method of claim 17, further comprising computing a confidence score associated with at least one of a likelihood that the event will occur or a likelihood that the event will be impactful.

### Claim 20 (depends on 19)

The method of claim 19, further comprising controlling the generating of the scaling strategy based on the confidence score.
